Pharmacoeconomic studies and the pharmaceutical industry: Bias in which direction?

Abstract

In summary, I believe that pharmacoeconomics and outcomes research are scientific disciplines, as such they should be based on good evidence. I have not seen evidence that there is systematic bias in economic and outcomes studies sponsored by the pharmaceutical industry. Furthermore, while steps have been taken to identify (and ideally control) potential biases in pharmaceutical industry-sponsored economic and outcomes research publications, a bias against industry-sponsored studies remains. Studies funded by the pharmaceutical industry are often assumed to be guilty until proven innocent, with few opportunities for a fair hearing. As discussed by Drummond, 'Bias is still perceived to be a major problem in industry-sponsored studies, both by healthcare decision-makers and journal editors,' despite the lack of evidence for biased study methodology [9]. Thus, I simply suggest we all play fair. Rather than assuming inherent flaws in studies funded by pharmaceutical companies, pharmacoeconomic and outcomes research should be judged on its merits. These disciplines are now developed enough that reviewers in appropriate areas of expertise can be found to critique economic and outcomes submissions. If a submission does not use good methodologies or does not provide sufficient information for complete evaluation, draws unwarranted conclusions from the results, it should be rejected regardless of the identity of the sponsor. In contrast, if a submission is good and has objective research, publication should ensue. In addition, if employees of a sponsoring company meet publication criteria for study authorship, they should not be denied the opportunity to be authors solely based on their employer. The sponsor's identity and relationship to the authors should be included with the publication, to avoid any appearance of bias. However, standards for authorship and manuscript review should not be based on concerns for this unproven bias and should be independent of the study's sponsor.